Improved Expectation Maximization Algorithm for Gaussian Mixed Model Using the Kernel Method

Table 4

Simulation results for the case 𝜇₁=0, 𝜇₂=0.25, 𝜎²₁=0.577, and 𝜎²₂=1.414.

Prior prob.		Bias, 𝐷𝑖
		𝑎1		𝑎2		𝜇1		𝜇2		𝜎^21		𝜎^22
𝑎1	𝑎2	𝐷	𝑆𝐷	𝐷	𝑆𝐷	𝐷	𝑆𝐷	𝐷	𝑆𝐷	𝐷	𝑆𝐷	𝐷	𝑆𝐷
0.1	0.9	0.089	0.007	−0.089	0.007	−0.817	3.770	0.045	0.048	−0.341	0.823	0.305	0.095
0.2	0.8	0.157	0.108	−0.157	0.108	0.291	3.521	0.075	0.073	−0.169	0.374	0.413	0.256
0.3	0.7	0.237	0.100	−0.237	0.100	−0.205	2.924	0.105	0.121	−0.278	0.389	0.578	0.278
0.4	0.6	0.245	0.187	−0.245	0.187	0.602	2.520	0.092	0.108	−0.008	0.258	0.508	0.435